Product Thinking for Data

By Dr. Chris Harding, Founder and Principal, Lacibus, Ltd, and Dr. Anas Tawileh, Senior Advisory Consultant, AWS


The role of data as a strategic enabler has become increasingly crucial to the long-term competitiveness of organisations. Netflix, for example, is one of the prominent examples of organisations that have successfully leveraged data to strengthen and sustain their competitive market position. One of the company’s many data-driven innovations is its market-leading Recommendations System that uses machine learning technologies to provide personalised recommendations that Netflix viewers will most likely enjoy. This system is so successful that it is estimated to create more than $1 billion in business value per year according to Netflix’s Chief Product Officer Neil Hunt.

Similar examples abound of organisations that were able to successfully unleash the potential of their data assets to create significant business and customer value, and several studies seem to confirm this trend. A recent study published by McKinsey & Company found that data-driven companies outperform their competitors by 20%. In this sense, it can be argued that the innovative use of data to create and sustain competitive advantage is a true game changer. It will generously reward organisations that understand the potential, adopt the right mindset, and build the required capabilities to enable these innovations. 

As these trends accelerate and continue to skew the competitive balance in favour of data-driven organisations, business executives and strategists seek answers to many emerging questions: What do we need to do to be able to successfully transform into a data-driven organisation? What skills, capabilities and relationships should we develop to drive this transformation? How can we assess the potential of our data assets to create new business and customer value?

To succeed in transforming their organisations to become data-driven, executives must adopt a value-focused approach to the way they build their data assets, how they identify opportunities to leverage these data assets to create business and customer value, and how they innovate to transform data assets into value drivers. One way to become value-focused is to think of data as products.

Thinking of Data as Products

Data products are a hot topic right now. There are some exciting new data platforms to support them, and the vendors are generating plenty of media buzz. But using data products is not just a question of buying a new platform. It has big implications for your organisation’s culture, governance, value delivery, and team structure.

The starting point for the culture change is for everyone to think of data in terms of products. This is a big step. A hundred years ago, no one thought of anything in terms of products. Neil H McElroy is credited with inventing the concept of product management at Procter & Gamble in 1931. Since then, the advantages of this way of thinking, as a better way of giving people the material goods that they need, have become clear. Now we are applying this concept to data, but people often don’t naturally think this way about something that is not material.

Thinking of data as products encourages a wider perspective on the data asset throughout its full lifecycle, starting from the point of conception all the way towards retirement and decommissioning. It also unlocks access to an expansive repertoire of tools, methodologies and techniques that have been tested and proven to optimise value delivery.

Key benefits of applying product thinking for data assets include:

  • Places emphasis on the asset’s value from the end user’s perspective, enabling higher levels of customer centricity
  • Enables the inclusion of value realisation mechanisms in the product design and build, leading to increased visibility of realised value
  • Facilitates the management of the data asset’s quality through a clear fitness for purpose lens
  • Supports data-driven decision making for product related decisions, including investment, adaptation and retirement

The Imperatives of New Ways of WorkingFostering the Right Culture

The ways that producers and consumers should work are well understood for material products. Producers should learn to understand their consumers, understand their needs, and do their best to meet those needs. Consumers should make their needs clear and give producers honest feedback. Data producers and consumers must be educated to develop these ways of working, and to be able to practice them effectively.

As an inspirational president, John F Kennedy once said, “Change is the law of life, and those who look only to the past or present are certain to miss the future.” As a practising politician, he may also have been aware of the advice given by Niccolo Machiavelli that, “There is nothing more difficult to take in hand, more perilous to conduct, or more uncertain in its success, than to take the lead in the introduction of a new order of things.”

You must clearly articulate the concepts of data products and data value. Do your people buy into them? Probably not at first. Perhaps they will pay “lip service.” Communication – both ways – is key to understanding and overcoming the reservations that you will encounter. Bringing the leaders on board is a crucial first step, but you must be sure to make your priorities clear to everyone. Culture change may be a perilous undertaking but, if you do not take the difficulties lightly, you can succeed.

Establishing the Right Governance Framework

New attitudes and ways of working imply new forms of governance. Data products governance is needed to settle questions such as who provides which data, who can consume which data, and what are the data products? To settle these questions, an organisation has policies and mechanisms for their interpretation and enforcement. Example policies might include, “Departments that collect data should provide it to others as appropriate and needed,” “Personally identifiable data should only be provided to consumers that can justify a need for it,” and “A data product should be defined when there is a set of related data that is needed by departments other than those that collect it.”

These policies will not be appropriate for every organisation; each organisation must decide what policies it needs. These policies clearly require interpretation. What does “appropriate and needed” mean? How can a consumer justify a need for personally identifiable data? What is a set of related data? Management roles must be identified, or governance boards established, to answer such questions. Management processes must be defined for situations where these roles or boards determine that the policies are not followed.

Creating the Right Data Products

Successful producers create data products that respond to user needs and provide significant user value. In these new ways of working, user-centric approaches to product design are effectively deployed to design and develop data products. Examples of such user-centric approaches include empathy mapping and Clayton Christensen’s “jobs to be done” framework. The latter is particularly useful as it focuses on identifying the real problem the user is trying to solve and developing products that help them solve this problem. Applying this framework to data assets helps in mapping the existing or potential data products to real customer needs in the marketplace and facilitates the product design to specifically satisfy these needs.

Several companies have successfully used the jobs to be done framework to leverage data and analytics to satisfy their customer needs. For example, Netflix identified a key need (or job to be done) for their customers: Customers wanted to be able to quickly find video content that they can enjoy. Once the architects and developers at Netflix identified this job, they started working together on creating innovative ways to satisfy it. The result was Netflix’s Recommendations System, a data product that used machine learning technologies to analyse the large amounts of data generated through Netflix’s customer use of the platform to offer viewers personalised recommendations that they will most probably enjoy. This thoughtfully designed data product served its intended “job to be done”, and resulted in increased viewer engagement, and directly contributed to Netflix’s bottom line.

The jobs to be done framework is a useful approach to product management that focuses the product design and development on the specific customer needs. It achieves this goal by offering a structured method to identify and communicate the customer’s jobs to be done, and to understand the thought process the customer goes through to “hire” this product. In the example of Netflix’s Recommendation System, the job the customer wants to get done was finding video entertainment they would enjoy watching without spending too much time in research. The thought process may look something like: “Netflix should be able to know what kind of content I might enjoy. When I am looking for something new to watch, I want easily accessible recommendations”.

Building the Right Team

There must also be changes to the organisation’s team structure. Product teams, new governance boards, and an architecture team will be needed. A product team could correspond to a business domain, or could be a cross-functional team with business domain and data specialist members.

The role of a data product manager is now well established. There are data product management training courses. Data product manager positions are frequently advertised, with an average salary, according to Zippia, of $120,000 in the US. The data product manager is entrusted with key decisions that impact the shape, plans and performance of the product, and is accountable for the product’s performance and benefit realisation.

The data product manager takes ownership and responsibility for the data product throughout its full lifecycle. She starts by identifying the user needs, mapping these needs to relevant data assets, capturing and developing the product’s functional and non-functional requirements, coordinating product delivery, monitoring and reporting performance, and making product end of life decisions.

In this role, the data product manager needs to collaborate with many stakeholders, including data custodians, governance and quality management, product designers, developers, and most importantly, users. A manager may own and manage a single data product, or may manage several data products, perhaps with data having several different owners.

As well as managers, the data product teams will need data specialists, and specialists in the technology of data product delivery. This no longer just means database administrators and programmers. In a world that is increasingly automated and cloud-native, they will need data-ops and cloud specialists.

Last, but not least, the organisation will need an architecture team. It will need Enterprise and IT architects to help define the data product principles, set technology and interface standards, design the technical solutions to support data products and their use, and monitor solution implementation. This includes not only the data platforms and storage systems. It should also include systems to support the identification and management of data products, for example by collecting statistics of data use, and data product consumer feedback.


Innovative use of data to create and sustain competitive advantage is driving corporate success. Data products are emerging as significant game changers. Organisations that are successfully packaging their data assets as valuable data products are reaping considerable business returns.

Thinking of data as products is not easy. Defining the right products requires a focus on customer value, and the ability to understand that value and measure it. Executives should encourage their teams to do this by leveraging proven user-centric approaches and methodologies. Delivering and managing the products needs a new team structure, including the key role of the data product manager and how this role will interact with the rest of the team. New governance processes are needed for smooth and effective data product team operations. Above all, the organisation must have a customer focused, data-driven culture.

None of this is easy, but it is all certainly worth doing. Product thinking revolutionised the design and delivery of material goods, and it will do the same for data. It is a discipline that helps companies deliver more value to their customers. And the companies that do this best are – deservedly – the ones that will continue to succeed.

Dr. Anas Tawileh is a Senior Advisory Consultant at AWS. He leads the strategy advisory practice in Canada, focusing on strategy, digital transformation, and business and technology architecture.

Dr. Tawileh has over twenty years in global experience advising senior executives and the C-suite on how to leverage technology to strengthen competitiveness, build differentiating capabilities, accelerate innovation, and realize business value. He has a PhD in Strategic Information Systems. 

Dr. Chris Harding is Founder and Principal of Lacibus Ltd. He formed the company to provide services based on virtual data lakes and data-centered architecture. Chris developed the ideas that led to the formation of the company while working as Director of the Open Platform 3.0 Forum of The Open Group.

Chris was a staff member of The Open Group for many years, supporting its members’ activities in data communications, directory interoperability, the Web, service-oriented architecture, cloud computing, and other areas.

He was the lead author of The Open Group Guide: Cloud Computing for Business, has helped produce a number of other Open Group publications, and has written many on-line articles. He remembers the early development of TOGAF(R) and maintains an interest in enterprise architecture as a member of the The Open Group Data Integration Work Group.

His main focus is now on data platforms, and particularly on their use in natural language processing applications. Before joining The Open Group, Chris was a data communications consultant and, before that, a software engineer and team leader. He has a PhD in Mathematical Logic. He lives with his wife in Lincolnshire, UK, where he has scope to pursue his hobbies of gardening and photography.